Physically based rendering in immersive virtual reality: an optimization approach for tuning lighting, color, and material parameters

Physically based rendering and immersive virtual reality are emerging as powerful tools for lighting design and visual evaluation. However, existing simulation tools often fail to reproduce the photometric and colorimetric characteristics of real-world environments when visualized in virtual reality. Accurate modeling of lighting and material behavior in Immersive Virtual Reality requires methodologies that bridge physical measurements with rendered visual output. Despite recent progress, methods for aligning real and virtual scenes photometrically and colorimetrically remain limited. This work presents a novel methodology that integrates spectral and photometric measurements with perceptual metrics and optimization algorithms to calibrate parameters of lighting and materials in physically-based rendered environments. We introduce two algorithms: one to compute an optimized RGB transformation matrix derived from spectral reflectance data, and another to adjust lighting and material properties by minimizing a composite perceptual metric that measures both colour and image structure. The results show that this approach enables automatic estimation of the physically-based rendering properties of light sources and materials. It also improves the colour space representation for virtual reality displays. A perceptual study involving 23 participants demonstrated high agreement between the appearance of real and virtual colours and produced strong scores on visual realism (<i>Q</i>₂ = 4.0/5.0) and satisfaction (<i>Q</i>₁₆ = 4.8/5.0). These findings demonstrate that integrating Physically Based Rendering with a perceptual metric enables the generation of photometrically and visually realistic scenes in Immersive Virtual Reality environments.